Abstract:

To solve a bottleneck problem of authoring a finite state automaton (FSA)-based ICALL system capable of automatically correcting free format English composition sentences, we have developed a new template-template scheme to simplify and streamline the labor-intensive input processes of the network of possible answers. This is achieved by exploiting a simple rule-based approach to the template construction which allows us to integrate complex template patterns into a simpler single template and also to generate many patterns, particularly those of grammatical errors, automatically. This approach contributes to a reduction both in computational processing time and complexity of the system and opens a wide door for a new natural language processing (NLP)-based dialogue system (written and spoken) of future man-machine interactive systems.

Author: Liang Chen, Naoyuki Tokuda